F.V.Vier & J. A. P. Figueiredo
Post-Graduate Programof Dentistry, ULBRA, Canoas, Brazil.

Introduction.
Following pulp necrosis, the root-canal system encourages the colonization and proliferation of microbes (Estrela & Figueiredo 1999). The low intensity, chronic stimulus provided by bacteria and their products allows for the maintenance of inflammation in the periapical region (Yanagisawa 1980, Lin et al.1984). Chronic periapical lesions bearinga proliferative character, represented by granulomas and periapical cysts, are the result of this process (Leonardo et al. 1998). The slow growth of these lesions results in bone resorption that is visible radiographically.
The precise nature of such lesions can only be determined histologically (Linenberg et al. 1964). However, true prevalence of each pathological condition is unclear. Cystic lesions have been reported to account for between 3.2% (Nair 1987) and 54% (Priebe et al. 1954) of apical lesions, and granulomas for between 45% (Lalonde & Luebke 1968) and 96.8% (Nair 1987). Some of this may be explained by differences in sample source and histological methods.
A number of chemical mediators of inflammation, including the cytokynes IL-1a, IL-1b, TNFa, prostaglandins and LPS, seem to be related to the pathogenesis of periapical lesions (Schein & Schilder 1975, Schonfeld et al. 1982, Burchett et al. 1988, Wang & Stashenko 1993a,b). These substances may stimulate root resorption in the same way that they stimulate bone resorption (Hammarstrom & Lindkog1992).
Although radiographic examination is an important resource in clinical diagnosis, it is rarely helpful in the diagnosis of small areas of external root resorption associated with teeth having apical periodontitis (Bhaskar & Rappaport1971, Ferlini Filho1999, Laux et al. 2000).
Irregular resorbed areas are frequently situated in sites that are not within the reach of root-canal instruments or medication and may act as niches for extraradicular bacterial colonization (Tronstad et al. 1990, Lomc¸ali et al.1996), besides causing technical problems for rootcanal treatment (Delzangles1989, Malueg et al.1996).
There are few reports correlating periapical lesions with external inflammatory root resorption (Delzangles 1989, Bohne 1990,Vier & Figueiredo 2000). The aim of this study was to evaluate the prevalence of different periapical lesions as well as to evaluate the presence and extension of apical external resorption and its association with different categories of disease.

Materials and methods.
The sample comprised 113 extracted human teeth with visible periapical lesions. These extracted teeth were obtained from the Public Dental Services in the state of Rio Grande do Sul, Brazil, where endodontic or restorative treatment is not provided. Some of the teeth were symptomatic, others were removed because the crown was completely compromised by dental caries. Care was taken to include only teeth with whole and intact periapical lesions.
The specimens were stored at room temperature in a solution of formaldehyde (10%, w/v), and radiographed to exclude previous root-canal treatment or incomplete root formation. Fifteen teeth were discarded, leaving a final sample of 98 teeth, amongst which six presented two lesions on separate roots. A total of 104 apical lesions and104 root apices were analyzed.
The periapical lesions were gently removed from the root apices and labeled before processing. Semi-serial sections (0.5 m at 0.5-mm intervals) of soft periapical lesions were stained with HE. The periapical lesions were classified as:

Noncystic lesions

• Periapical granuloma: Lesions predominantly in filtrated by lymphocytes, plasma cells and macrophages, with or without epithelial remnants (Nair et al.1996), and covered by a capsule of collagen fibres. In these lesions, neutrophils were sparse forming no abscess microcavities or concentrated infiltrates.
• Periapical abscess: Lesions with a distinct collection of neutrophils in the interior of a previously existent granuloma (Nair et al.1996). These were further categorized as 1 јabscess cavity occupying up to one third; 2 ј1/3-2/3; 3 _2/3 of the total area of the visualized lesion in the histological sections.

Cystic lesions

• Periapical cyst: Lesions with a layer of stratified squamous epithelium a long a surface of sufficient quantity of conjunctive tissue to indicate a delineated cavity and surrounded by a slight fibrous capsule (Patterson et al.1964, Lalonde&Luebke1968, Lalonde1970,Nobuhara & Del Rio1993,White et al.1994).
• Abscessed periapical cyst: Cysts containing pus-filled cavities were classed as1, 2 or 3 as above.

The histological sections were analysed by two blinded and previously calibrated observers, using 32, 100 and 400x magnification (Zeiss microscope, Thornwood, USA).
The diagnosis of each lesion was determined by considering all the histological sections. The presence of epithelium delineating a pathological cavity in one or more sections of a lesion characterized it as a periapical cyst. Classification of abscess severity was based on the section showing its largest dimension. Thereafter, the periapical lesions were grouped in two subgroups:

• noncystic lesions: A1 - with absence or with a small abscess (periapical granuloma and abscess degree 1); A2 - with advanced abscess (periapical abscess degrees 2 and 3).
• cystic lesions: B1-withabsence or with a small abscess (periapical cyst and abscessed cyst degree1); B2-with advanced abscess (abscessed cysts degree 2 and 3).

The apical portion of the root was cut perpendicular to the long axis with carborundum disc to favour placement on stubs. The root apices were submerged in a solution of 2.5% sodium hypochlorite for 3 h, dehydrated in an ascending sequence of alcohols (70, 90 and 99.96%, for 5 h in each), attached to the stubs with the apex up wards, and sputtered (Balzers, Liechtensten) with gold palladium, to a thickness of 150 A. Scanning electron microscopy was conducted using a PhilipsXL 20 (Eindhmoven, Netherlands) microscope, operating at 15 Kv. The areas that surrounded the apical foramina of the apices were imaged at 100_. The apical foramen occupied the central region of the video screen and was totally surrounded by the root. If a root presented more than one apical foramen, more than one image was obtained, so as to guarantee the analyses of all the foramina. SEM images were analysed separately by two blinded and previously calibrated observers.
The dental apices were classified depending on the presence or absence of external apical resorption, as well as to its extent:

Periforaminal resorption:
Periforaminal resorption was defined as the area of resorption not comprising the outline of the foramen, but the surrounding area. The degrees of severity 0-3 were employed when there was absence of resorption, resorption of up to 1/4, from 1/4 to1/2 and in more than1/2 of the area that surrounded or circumscribed the apical foramen, respectively.

Foraminal resorption:
Defined as the resorption within the outline or the perimeter of the foramen. The degrees of severity 0-3 were employed when there was absence of resorption, resorption of up to 1/4, from 1/4 to 1/2 and in more than1/2 of the area of the outline or the perimeter of the apical foramen, respectively.
Whenever a tooth presented with two or more apical foramina or had fused roots, the degree of final resorption measured was the foramen where the resorption was most severe.
Kappa coefficient was employed to evaluate the degree of agreement amongst the examiners, both in diagnosing the periapical lesions and in the presence and extension of periforaminal and foraminal external resorption.
The anova statistical test was used to evaluate the correlation between the histopathological diagnosis of the periapical lesions and the presence and extension of the external apical root resorption, followed by the Duncan test of multiple comparisons. Due to the ordinal nature of the data, the Kruskal-Wallis anova was also used. The frequencies observed in the groups were compared using the Chi-square test.

Histopathology.
The results are presented in Tables 1-5. Analysis was completed in a total of 102 specimens, as two were lost during preparation. The degree of agreement amongst the observers concerning the diagnosis of the periapical lesions gave a Kappa coefficient of 0.96 (IC 95% ј 0.82-1.0).
Periapical granulomas was not a common finding. A1 lesions (granuloma and periapical abscess score1) comprised only11.8%of the sample (Table 1).The most prevalent histological diagnosis was the lesions classified as A2 (noncystic periapical abscess scores 2 and 3) (Fig.1), which comprised 63.7% of the sample (Table 1).
Periapical cysts represented 24.5% of the sample. Just four (3.9%) periapical cysts presented with no or small abscess cavities (Fig. 2). Amongst the 25 cysts analysed, 21 (84%) had large abscess cavities. The abscessed areas were associated with an epithelial component resulting in disorganization and discontinuity. The greater the severity of the abscess, the bigger the epithelial disintegration observed. In other cases, the cavities of the abscesses were situated mainly on the connective tissue adjacent to the epithelial layer (Fig. 3).
There were cases where focal areas of neutrophils were absent within the cyst epithelium in one histological section, but the following section often revealed a defined abscess in that site. Disregarding the cystic characteristic of the lesions and by just analysing the presence of an abscess, we were able to observe that 86 lesions (84.3%) had abscess cavities that occupied a large area. Just 16 lesions (15.7%) were free from or contained only microcavities of acute inflammatory cells (Table 2).

Table 1. Prevalence of different periapical lesions.



Table 2. Periapical lesions of minor and major severity of abscess.


A1-noncystic lesions with absence or with a small degree of abscess.
A2-noncystic lesions with high degree of abscess.
B1-cystic lesions with absence or with a minimum degree of abscess.
B2-cystic lesions with high degree of abscess.

Resorption.
Only 12.7% of the samples were free of periforaminal resorption and 16.8% showed integrity of the foraminal surroundings (Table 3). From all the samples, 72.6% had periforaminal resorption, reaching more than 1/ 4th of the area around the foramen (Fig. 4). Moreover, in 42.2% of these cases, the resorption included more than half of this area (Fig. 5). When foraminal resorption occurred, 58.4% of the samples were affected in more than 1/4 of their perimeter (Fig. 6), and 28.7% in more than half (Fig.7). Just 8.9%of the samples had no resorption (Table 4).
From Table 4, it is clear that the pattern of periforaminal resorption did not depend on the pattern of foraminal resorption or vice-versa. An apex may present with substantial periforaminal resorption, without alteration to the foramen. The apical foramen may be resorbed, even if the zone that surrounds it had an intact cementum structure. However, the resorption could occur both in a periforaminal and foraminal location and to different degrees.
By associating the presence of periforaminal and foraminal resorption, we were able to perceive that 49.5% of the apices had resorption in more than 1/4th of the examined area, and in more than half in 22.8% of cases (Table 4) (Fig.7). The lacunae of resorption were similar to the Howship type, showing a circular shape of different sizes. The combination of various lacunae resembled the characteristic aspect of honeycombs (Fig. 5). Generally, the areas of resorption were superimposed, projecting pronounced margins (Fig. 4). However, in some cases, due to the fusion of the lacuna, or due to its shallowness, the interlacunae crests became less obvious or even absent (Figs 4 and 6).

Table 3. Extension degrees of periforaminal and foraminal resorption of the dental apexes.


* In one specimen, it was not possible to performa classification of the foraminal resorption.

Table 4. Distribution of degrees of periforaminal and foraminal resorption combined.



Table 5. Median (p25-75) of the degrees of periforaminal and foraminal resorption according to the different periapical lesions.


Periforaminal: P = 0.227; foraminal: P = 0.163.
A1-noncystic lesions with absence or with a small degree of abscess.
A2-noncystic lesions with high degree of abscess.
B1-cystic lesions with absence or with a minimum degree of abscess.
B2-cystic lesions with high degree of abscess.
* In one specimen, it was not possible to perform a classification of the foraminal resorption.

The resorbed areas resulted in the presence of one or two isolated lacunae, surrounded by an integral cementum surface or by a set of interconnected gaps that were occasionally related to other resorption zones. Sometimes, the apex had fused areas of resorption, and was totally deprived of intact cementum (Fig. 5). There were cases with evidence of exposed dental tubules (Fig. 8), demonstrating sites of increased resorption.
The foraminal resorption viewed by SEM followed two peculiar characteristics. In some apices, there was deformation of the original foramen outline, due to deep resorption, though in very specific places (Fig. 9). In others, although the apical foramen had not suffered visible morphological alterations, it was possible to observe shallow resorption gaps, surrounding the majority of its outline (Figs 5 and 8).
According to Table 5, we can verify that there was no statistically meaningful difference between the type of periapical lesion and the degree of periforaminal and foraminal resorption at the root apex.

Figure 1. Periapical abscess degree 3 (14x).



Figure 2. Periapical cyst (cystic lesion with the absence of abscess microcavities) (32x).



Figure 3. Abscessed cyst degree 3, with abscessed area adjacent to the epithelial component (14x).



Figure 4. Apex with periforaminal resorption involving1/4-1/2 of the area. Pronounced areas of lacunae (A). Interlacunae crests were not much evident (B) (PR:2; FR:2; lesion: B2) (100x).



Figure 5. Apex with periforaminal resorption involving more than half of the radicular surface examined. The resorption displayed a honeycomb aspect (PR:3; FR:3; lesion: A2) (100x).



Figure 6. Apex with foraminal resorption involving1/4-1/2 of the perimeter with a not well defined, or even absent outline of the interlacunae crests (PR:1; FR:2; lesion: B2) (100x).



Figure 7. Apex with periforaminal and foraminal resorption involving more than half of the radicular surface examined (PR:3; FR:3; lesion: A2) (100x).



Figure 8. Apex with extensive periforaminal and foraminal resorption exposing dentinal tubules (PR:3; FR:2; lesion: A2) (100x).



Figure 9. Apex with foraminal resorption with morphological alterations of the apical contour (PR:3; FR:3; lesion: A2) (100x).

Discussion.
Many teeth with apical periodontitis are believed to show some degree of external periapical resorption, but few studies have related resorption with the nature of the apical lesion (Delzangles 1989, Bohne 1990, Vier & Figueiredo 2000). This study sought to address this issue by correlating periapical pathology with the presence and extension of periforaminal and foraminal resorption.
The presence of periapical pathosis was determined by direct visualization of root apices (Linenberg et al. 1964, Simon1980, Garrocho & Antonio Neto 1984, Nair 1987, Bohne1990, Nair et al.1996). Radiographic examination was not conducted prior to tooth extraction. The presence of a periapical radiolucency does not necessarily indicate periapical inflammation, other causes may include apical scar tissue (Nair et al. 1999). Conversely, a periapical inflammatory reaction of endodontic origin may exist without being visible radiographically (Bender & Seltzer1961).
Semi-serial sections were performed on the specimens, since abscess cavities or even the epithelial lumen of a cystic cavitymay be away from the centre of the specimen or from the place elected for randomized sections (Garrocho & Antonio Neto 1984), a fact which could make it difficult to reach the final diagnosis. The most accurate analysis is achieved from serial sectioning of intact lesions.
We classified abscesses as lesions without epithelium in the body of the lesion and with a massive accumulation of neutrophils. The lesions diagnosed as periapical granulomas exhibited neutrophils, though not forming distinct arrangements or fina degenerating state. We classified cysts according to Patterson et al. (1964), Lalonde & Luebke (1968), Lalonde (1970), Nobuhara & del Rio (1993) and White et al. (1994), because when cysts become infected the epithelium may not be complete or intact, but disintegrated in some areas.
After a thorough hand individual analysis of each histological section, we agreed with Linenberg et al. (1964), who demonstrated transitional stages between one type of lesion and another when analysing various histological sections of the same lesion.
Differences in histological criteria create difficulty in comparing similar studies (Langeland et al.1977, Spatafore et al.1990, Nobuhara & del Rio1993). Our study, for example, demonstrated an unusually low prevalence of periapical granuloma and high prevalence of noncystic abscesses. This may be due to the sample, which consisted of teeth extracted in the Public Oral Services in poor areas in southern Brazil, where the presence of symptoms is one of the main reasons for dental visits.
The prevalence of 24.5% of cystic lesions is in accord with previous reports (Linenberg et al.1964, Langeland et al. 1977, Nobuhara & del Rio 1993, White et al. 1994). However, when evaluating other studies with similar methodology, i.e. lesions associated with the apices of extracted teeth, the results were higher than the 17 and15%of the cysts diagnosed by Simon (1980) and Nair et al. (1996) respectively. The results were closer to the 28% reported by Linenberg et al. (1964), whose samples included material curetted from extraction sockets and lesions associated with extracted teeth.
This study also evaluated the extension of resorption over the apical root surface. The results are difficult to correlate with many others who simply reported the presence of resorption on the root surface (Simon et al. 1981, Hess et al.1983, Laurent-Maquin et al.1986,Delzangles 1989, Bohne 1990, Lomc¸ ali et al. 1996, Bonifa. cio et al. 2000).
From the methodology employed, 8.9% of specimens had no periforaminal or foraminal resorption, despite pulp necrosis and periapical lesions. These results are in agreement with those of Henry & Weinmann (1951) and Ferlini Filho (1999) who reported that 10% of their specimens had no resorption.
Apical periodontitis is amongst the causes of progressive inflammatory resorption cited by Soares & Goldberg (2001). Nevertheless, it is impossible to determine the influence of other factors that might potentiate resorption such as occlusal trauma. We had no access to detailed medical or clinical data to control for such parameters.
The apical resorption observed in the present study did not differ amongst the various periapical pathological conditions, since all of them exhibited a common feature, i.e. the presence of inflammation. This is an important aspect to consider in the histological description of these cystic and noncystic lesions, independent of the degree of abscess. Both constitute a continuous and variable aspect of the same phenomenon, inflammation. The physical presence of a periapical lesion could promote root resorption in addition to bone lysis, as it makes room for its organization and growth, independently of its histological classification. Thus, the results of the present study differ from those of Delzangles (1989) and Vier & Figueiredo (2000), who included a small number of cystic samples and used less refined histological sectioning.
Normally, the failure of root-canal treatment is related to the persistence of infection in the root-canal system (Sjogren et al. 1997, Sundqvist et al. 1998). The presence of root resorption in teeth with periapical lesions is important for infection control since these areas are niches for bacteria. Moreover, the apical limit of instrumentation may be altered in teeth with widely resorbed apices, since the cementum-dentine junction at the constriction can be missing (Delzangles 1989, Malueg et al. 1996). In such circumstances, sealing the canal may be difficult and over filling is likely.
Bacteria may also be found in the external surface of the root, forming a periapical biofilm (Leonardo & Silva 1998), thus confirming the importance of appreciating the existence of periforaminal root resorption. No bacterial bio films were, however, demonstrated in this study.

References.

Bender I, Seltzer S (1961) Roentgenographic and direct observation   of experimental lesions in bone: I. Journal of the American Dental Association   62, 152-60.
Bhaskar S, Rappaport H (1971) Histologic evaluation of endodontic procedures   in dogs. Oral Surgery, Oral Medicine and Oral Pathology 31,526-35.
Bohne W (1990) Light and ultrastructural studies of human chronic periapical   lesions. Journal of Oral Pathology andMedicine19, 215-20.
Burchett S, Weaver W, Westall J, Larsen A, Kronheim S, Wilson C (1988) Regulation   of tumor necrosis factor/cachectin and IL-1 secretion in human mononuclear phagocytes. Journal of Immunology140, 3473-81.
Delzangles B (1989) Scanning electronmicroscopic study of apical and intracanal   resorption. Journal of Endodontics 15, 281-5.
Lindkog S (1992) Factors regulating and modifying dental root resorption. Proceedings of the Finnland Dental Society 88 (Suppl.1), 115-23.
Henry J, Weinmann J (1951) The pattern of resorption and repair of human   cementum. Journal of the American Dental Association 42, 270-90.
Hess J, Culieras M, Lamiable N(1983) A scanning electron microscopic investigation   of principal and accessory foramina on the root surfaces of human teeth: thoughts   about endodontic pathology and therapeutics. Journal of Endodontics 9, 275-81.
Lalonde E (1970) A new rationale for the management of periapical granulomas   and cysts: an evaluation of histhopathological and radiographic findings. Journal of the American Dental Association 80, 1056-9.
Lalonde E, Luebke R (1968) The frequency and distribution of periapical cysts   and granulomas. Oral Surgery, Oral Medicine, Oral Pathology 25,861-8.
Langeland K, Block R, Grossman L (1977) A histopathologic and histobacteriologic   study of 35 periapical endodontic surgical specimens. Journal of Endodontics   3, 8-23.
Laurent-Maquin D, Blocquaux Verchere M, Bouthors S (1986) Modifications morphologiques   radiculaires de dents atteintes d'une le. sion pe. riapicale. Etude au microscope   electronique a' balayage. Revue d'Odonto-Stomatologie 15,379-84.
Laux M, Abbott P, Pajarola G, Nair P (2000) Apical inflammatory root resorption:   a correlative radiographic and histological assessment. International Endodontic   Journal 33, 483-93.
Leonardo M, Leal J, Lia R, Martins J (1998) Filosofia do tratamento de canais   radiculares Necropulpectomia: Conceituac. In: Leonardo M, Leal J, eds. Endodontia:   Tratamento de Canais Radiculares, 3rd edn. Sao Paulo, Brazil: Panamericana,   pp.127-58.
LeonardoM, Silva L (1998) Medicac topica entre sessoes, 'curativo de demora'   em biopulpectomia e necropulpectomia I e II. In: LeonardoM, LealJ, eds.   Endodontia: Tratamento de Canais Radiculares, 3rd edn. Sao Paulo, Brazil: Panamericana,   pp. 491-534.
Lin L, Shovlin F, Skribner J, Langeland K (1984) Pulp biopsies from the teeth   associated with periapical radiolucency. Journal of Endodontics 10, 436-48.
Linenberg W, Waldron C, Delaune G Jr (1964) A clinical roentgenographic and   histhopathologic evaluation of periapical lesions. Oral Surgery, Oral Medicine   and Oral Pathology 17, 467-72.
Lomcali G, Sen B, Cancaya H (1996) Scanning electron microscopic observations   of apical root surfaces of teeth with apical periodontitis. Endodontics   and Dental Traumatolology 12,70-6.
Malueg L, Wilcox L, Johnson W (1996) Examination of external apical root   resorption with scanning electron microscopy. Oral Surgery, Oral Medicine,   Oral Pathology and Oral Radiology, Endodontics 82, 89-93.
Nair P (1987) Light and electron microscopic studies of root canal flora   and periapical lesions. Journal of Endodontics 13, 29-39.
Nair P, Pajarola G, Schroeder H (1996) Types and incidence of human periapical   lesions obtained with extracted teeth. Oral Surgery, Oral Medicine and Oral   Pathology 81, 93-102.
Nair P, Sjogren U, Figdor D, Sundqvist G (1999) Persistent periapical radiolucencies   of root-filled human teeth, failed endodontic treatments, and periapical scars.   Oral Surgery, Oral Medicine, Oral Pathology and Oral Radiology, Endodontics   87, 617-27.
Nobuhara W, Del Rio C (1993) Incidence of periroot pathoses in endodontic   treatment failures. Journal of Endodontics 19, 315-8.
Patterson S, Shafer W, Healey H (1964) Periapical lesions associated with   endodontically treated teeth. Journal of the American Dental Association   68, 191-4.
Priebe W, Lazansky J, Wuehrmann A (1954) The value of the roentgenographic   film in the differential diagnosis of periapical lesions. Oral Surgery,   Oral Medicine and Oral Pathology 7, 979-83.
chein B, Schilder H (1975) Endotoxin content in endodontically involved   teeth. Journal of Endodontics1, 19-21.
Schonfeld S, Greening A, Glick D, Frank A, Simon J, Herles S   (1982) Endotoxic activity in periapical lesions. Oral Surgery, Oral Medicine   and Oral Pathology 53,82-7.
imon J (1980) Incidence of periapical cysts in relation to the root canal. Journal of Endodontics 6,845-8.
imon J, Yonemoto G, Bakland L (1981) Comparison of cellular cementum in   normal and diseased teeth - a scanning electron microscopic study. Journal   of Endodontics 7, 370-5.
jogren U, Figdor D, Persson S, Sundqvist G (1997) Influence of infection   at the time of root filling on the outcome of endodontic treatment of teeth   with apical periodontitis. International Endodontic Journal 30, 297-306.
oares I, Goldberg F (2001) Endodontia: Tecnica E Fundamentos, 1st edn. Porto Alegre, Rio Grande do Sul, Brazil: Artes Me. dicas Sul. pp. 339-68.
patafore C,Grin J, Keyes G, Wearden S, Skidmore A (1990) Periapical biopsy   report: an analysis over a10-years period. Journal of Endodontics16, 239-41.
undqvist G, Figdor D, Persson S, Sjogren U (1998) Microbiologic analysis   of teeth with failed endodontic treatment and the outcome of conservative re-treatment.   Oral Surgery, Oral Medicine and Oral Pathology 85,86-93.
Tronstad L, Barnett J, Cervone F (1990) Periapical bacterial plaque in teeth   refractory to endodontic treatment. Endodontics and Dental Traumatolology   6, 73-7.
Vier F, Figueiredo J (2000) Morphologic analysis of apical resorption on   human teeth with periapical lesions. Endodontics Clinical Practice, Education   and Research 2(3). On-line publication of the University of Sao Paulo: http://www.usp.br  
  Wang C, Stashenko P (1993a) Characterization of bone-resorbing activity in   human periapical lesions. Journal of Endodontics19, 107- 11.
Wang C, Stashenko P (1993b) The role of interleukin_1alpha in the pathogenesis   of periapical bone destruction in a rat model system. Oral Microbiology   and Immunology 8, 50-6.
White S, Sapp P, Seto B, Mankovich N (1994) Absence of radiometric differentiation   between periapical cysts and granulomas. Oral Surgery, Oral Medicine and   Oral Pathology 78,650-4.
Yanagisawa S (1980) Pathologic study of periapical lesions 1. Periapical   granulomas: clinical, histopathologic and immunohistopathologic studies. Journal of Oral Pathology 9, 288-300.